http://franfabrizio.dev/tags/llms/ Recent content in llms on franfabrizio.dev Hugo en-us Fri, 01 May 2026 00:00:00 -0500 http://franfabrizio.dev/posts/adventures-in-local-llms-part-4/ Fri, 01 May 2026 00:00:00 -0500 http://franfabrizio.dev/posts/adventures-in-local-llms-part-4/ <h2 id="introduction">Introduction</h2> <p>In <a href="http://franfabrizio.dev/posts/adventures-in-local-llms-part-3/">Part 3</a>, I walked through the software stack I ended up running — Ollama, Open WebUI, paperless-gpt, and a handful of other tools — and how much of it felt like a science experiment. The stack worked, mostly, but it was only half the equation. The other half is the model itself.</p> <p>Once you&rsquo;ve got your hardware constraints figured out (we covered those exhaustively in <a href="http://franfabrizio.dev/posts/adventures-in-local-llms-part-2/">Part 2</a>) and your software stack in place, you&rsquo;re left with a deceptively simple question: <em>which model do you actually run?</em></p>